Mafia rating/ELO/reputation system?

This forum is for discussion related to the game.
User avatar
davesaz
davesaz
He
Survivor
User avatar
User avatar
davesaz
He
Survivor
Survivor
Posts: 12556
Joined: August 24, 2014
Pronoun: He
Location: Socially distant

Post Post #50 (ISO) » Fri Oct 06, 2017 4:05 am

Post by davesaz »

You'll say apathy does that too, but it's better to not join than it is to join, get a vanilla role, and replace out.
Wanna play Minecraft with your ms friends? Check out the minecraft thread, or the channel on discord
User avatar
Psyche
Psyche
he/they
Survivor
User avatar
User avatar
Psyche
he/they
Survivor
Survivor
Posts: 10048
Joined: April 28, 2011
Pronoun: he/they

Post Post #51 (ISO) » Fri Oct 06, 2017 4:48 am

Post by Psyche »

some other thread
youtube playlist extracter | donbot | game scraper | vca | setupsim | strategist | llm
User avatar
ironstove
ironstove
Mafia Scum
User avatar
User avatar
ironstove
Mafia Scum
Mafia Scum
Posts: 1742
Joined: July 22, 2016
Contact:

Post Post #52 (ISO) » Fri Oct 06, 2017 6:35 am

Post by ironstove »

Replacing out is bad unless it's a IRL issue. I think.

Replacing out causes the game to freeze and all other players to wait for a replacement, it causes the mod to have to do more leg work to find a replacement.

It is a negative action on the game, and it should definitely impact a player score for doing it. The replacement penalty will be small, so an occasional replacement should have no impact whereas a player who replaces often will have it reflected on his score.

I like some of the ideas presented. Rather than a behavior score, it would probably be a better idea to run an entire medley of stat tracking behind it that can be expanded out.

So far, stats I'm finding interesting, which won't necessarily impact the score but are just for your viewing pleasure:

-number of mislynches (lynches which you were on the train of which turned out to be incorrect)
-number of correct lynches
-number of times you have been mislynched (town stat)
-number of times you have been lynched (scum stat)
-replacing out
-replacing in
-total win/lose
-town win/lose
-scum win/lose
-3rd party win/lose

I would be interested in continuing the proj elo code, maybe some of it can be recycled as a stat tracker, but Realo I'd like to know how you were collecting the data for your ELO system, was it through self-reporting by the mods, or were you using some web crawler or API to parse the data off the forum?
User avatar
Realeo
Realeo
Jack of All Trades
User avatar
User avatar
Realeo
Jack of All Trades
Jack of All Trades
Posts: 5238
Joined: February 11, 2016
Location: Indonesia

Post Post #53 (ISO) » Fri Oct 06, 2017 6:37 am

Post by Realeo »

We copied the data from Normals and Newbies provided by listmod and parse it via some python script. Game likes Open who are not compiled by listmod are a pain in the ass.
"The debate on whether short multi postings or a long wall of post is good or not is like a debate on gun control--we would never understand each other and we have to make peace with it." -Realeo

I'm mabye a serious player, but I'm capable of joke. Ok?
User avatar
Thestatusquo
Thestatusquo
He/Him
Shea

User avatar
User avatar
Thestatusquo
He/Him
Shea

Shea

Posts: 14362
Joined: July 27, 2006
Pronoun: He/Him
Location: Chicago!

Post Post #54 (ISO) » Fri Oct 06, 2017 6:43 am

Post by Thestatusquo »

A player playing in a game they are not enjoying and do not want to be playing is far worse than a replacement.

If you have a game with 3-4 players who are not enjoying it and are feeling apathetic and such the game becomes miserable. I'd much rather those players just replace out.
tout comprendre c'est tout pardonner
User avatar
davesaz
davesaz
He
Survivor
User avatar
User avatar
davesaz
He
Survivor
Survivor
Posts: 12556
Joined: August 24, 2014
Pronoun: He
Location: Socially distant

Post Post #55 (ISO) » Fri Oct 06, 2017 6:52 am

Post by davesaz »

OK, there are situations where replacing out is acceptable.

What I'm trying to avoid is a player signing up for everything and shopping for games/roles they prefer by replacing out. I think the proposed stats could reveal that kind of behavior.
Wanna play Minecraft with your ms friends? Check out the minecraft thread, or the channel on discord
User avatar
Thestatusquo
Thestatusquo
He/Him
Shea

User avatar
User avatar
Thestatusquo
He/Him
Shea

Shea

Posts: 14362
Joined: July 27, 2006
Pronoun: He/Him
Location: Chicago!

Post Post #56 (ISO) » Fri Oct 06, 2017 7:03 am

Post by Thestatusquo »

Yeah, I agree that is bad.

I don't think its captured by your current system, and I know of at least one player who has been banned for doing that, so its not like its a big problem without a solution currently.
tout comprendre c'est tout pardonner
User avatar
Psyche
Psyche
he/they
Survivor
User avatar
User avatar
Psyche
he/they
Survivor
Survivor
Posts: 10048
Joined: April 28, 2011
Pronoun: he/they

Post Post #57 (ISO) » Fri Oct 06, 2017 7:23 am

Post by Psyche »

In post 54, Thestatusquo wrote:A player playing in a game they are not enjoying and do not want to be playing is far worse than a replacement.

If you have a game with 3-4 players who are not enjoying it and are feeling apathetic and such the game becomes miserable. I'd much rather those players just replace out.
if we're also tracking stats that would change if someone is staying in a game they're not enjoying being apathetic about and weight them accordingly, it should all even out then no?
youtube playlist extracter | donbot | game scraper | vca | setupsim | strategist | llm
User avatar
Thestatusquo
Thestatusquo
He/Him
Shea

User avatar
User avatar
Thestatusquo
He/Him
Shea

Shea

Posts: 14362
Joined: July 27, 2006
Pronoun: He/Him
Location: Chicago!

Post Post #58 (ISO) » Fri Oct 06, 2017 7:28 am

Post by Thestatusquo »

How would you capture that data?
tout comprendre c'est tout pardonner
User avatar
Psyche
Psyche
he/they
Survivor
User avatar
User avatar
Psyche
he/they
Survivor
Survivor
Posts: 10048
Joined: April 28, 2011
Pronoun: he/they

Post Post #59 (ISO) » Fri Oct 06, 2017 7:54 am

Post by Psyche »

well, you think they’d be playing worse in these cases, right?
they’d maybe be on fewer scum lynches, more mislynches, get mislynched more often, etc?
youtube playlist extracter | donbot | game scraper | vca | setupsim | strategist | llm
User avatar
Thestatusquo
Thestatusquo
He/Him
Shea

User avatar
User avatar
Thestatusquo
He/Him
Shea

Shea

Posts: 14362
Joined: July 27, 2006
Pronoun: He/Him
Location: Chicago!

Post Post #60 (ISO) » Fri Oct 06, 2017 8:15 am

Post by Thestatusquo »

I think thats speculative at best.
tout comprendre c'est tout pardonner
User avatar
ironstove
ironstove
Mafia Scum
User avatar
User avatar
ironstove
Mafia Scum
Mafia Scum
Posts: 1742
Joined: July 22, 2016
Contact:

Post Post #61 (ISO) » Sat Oct 07, 2017 2:59 am

Post by ironstove »

Yea, I agree you're trying to assess too much from the stats at that point. I think you might be able to add another layer like the win/lose of games you choose to replace out of, but that seems like such a pointless statistic.
User avatar
Psyche
Psyche
he/they
Survivor
User avatar
User avatar
Psyche
he/they
Survivor
Survivor
Posts: 10048
Joined: April 28, 2011
Pronoun: he/they

Post Post #62 (ISO) » Sat Oct 07, 2017 3:00 am

Post by Psyche »

In post 60, Thestatusquo wrote:I think thats speculative at best.
so you think player apathy only makes games less enjoyable?
there might actually be a way to measure how players impact game enjoyabity too through sentiment analysis...
youtube playlist extracter | donbot | game scraper | vca | setupsim | strategist | llm
User avatar
ironstove
ironstove
Mafia Scum
User avatar
User avatar
ironstove
Mafia Scum
Mafia Scum
Posts: 1742
Joined: July 22, 2016
Contact:

Post Post #63 (ISO) » Sat Oct 07, 2017 3:04 am

Post by ironstove »

So a few things I'm trying to figure out:

1. How to collect the (majority) of previous data accurately to record the stats I mentioned above that doesn't involve manually parsing through each thread or having someone else do the same.

2. How to collect future data accurately without the same issues as #1.

Trying to figure things out like if a player mislynched or correctly lynched seems like too difficult of a stat to collect, I think even just doing a basic win/lose record is hard. If data is standardized or structured, I can see myself easily compiling it all together, but having glanced through a few threads, that doesn't seem to be the case. It looks like every mod has their own way of reporting votes/scores/replacements and changes etc..

As for tracking replacements, I figured I could go through the queue/replacement thread to figure that out, but also seeing some inconsistencies that are going to cause headaches.
User avatar
ironstove
ironstove
Mafia Scum
User avatar
User avatar
ironstove
Mafia Scum
Mafia Scum
Posts: 1742
Joined: July 22, 2016
Contact:

Post Post #64 (ISO) » Sat Oct 07, 2017 3:04 am

Post by ironstove »

In post 62, Psyche wrote:
In post 60, Thestatusquo wrote:I think thats speculative at best.
so you think player apathy only makes games less enjoyable?
there might actually be a way to measure how players impact game enjoyabity too through sentiment analysis...
That sounds really interesting and all, but that's not what this thread is about, so maybe you could make your own topic about how certain statistics which people don't find relevant may actually provide insights which are overlooked.
User avatar
Psyche
Psyche
he/they
Survivor
User avatar
User avatar
Psyche
he/they
Survivor
Survivor
Posts: 10048
Joined: April 28, 2011
Pronoun: he/they

Post Post #65 (ISO) » Sat Oct 07, 2017 3:26 am

Post by Psyche »

yah sry it’s hard to focus sometimes
youtube playlist extracter | donbot | game scraper | vca | setupsim | strategist | llm
User avatar
Psyche
Psyche
he/they
Survivor
User avatar
User avatar
Psyche
he/they
Survivor
Survivor
Posts: 10048
Joined: April 28, 2011
Pronoun: he/they

Post Post #66 (ISO) » Sat Oct 07, 2017 3:30 am

Post by Psyche »

I have a plan for achieving 1/2 by expanding some code i’ve already produced and cleaning up our existing game archives, but it’ll take time to complete. In the meantime, if we can clarify details besides data collection per se like where/how we will represent the data, that his can be accellerated substantially.
youtube playlist extracter | donbot | game scraper | vca | setupsim | strategist | llm
User avatar
Mulch
Mulch
Gotta Go Fast
User avatar
User avatar
Mulch
Gotta Go Fast
Gotta Go Fast
Posts: 16940
Joined: April 23, 2017

Post Post #67 (ISO) » Sat Oct 07, 2017 10:08 am

Post by Mulch »

In post 5, Mulch wrote:This will have bad effects
Amrock#3784
User avatar
Psyche
Psyche
he/they
Survivor
User avatar
User avatar
Psyche
he/they
Survivor
Survivor
Posts: 10048
Joined: April 28, 2011
Pronoun: he/they

Post Post #68 (ISO) » Sat Oct 07, 2017 10:28 am

Post by Psyche »

and also good effects!
youtube playlist extracter | donbot | game scraper | vca | setupsim | strategist | llm
User avatar
Mulch
Mulch
Gotta Go Fast
User avatar
User avatar
Mulch
Gotta Go Fast
Gotta Go Fast
Posts: 16940
Joined: April 23, 2017

Post Post #69 (ISO) » Sat Oct 07, 2017 10:29 am

Post by Mulch »

In post 68, Psyche wrote:and also good effects!
Like what? See none that come without drawbacks
Amrock#3784
User avatar
Psyche
Psyche
he/they
Survivor
User avatar
User avatar
Psyche
he/they
Survivor
Survivor
Posts: 10048
Joined: April 28, 2011
Pronoun: he/they

Post Post #70 (ISO) » Sat Oct 07, 2017 10:31 am

Post by Psyche »

In post 69, Mulch wrote:See none that come without drawbacks
this will have bad effects and also good effects
youtube playlist extracter | donbot | game scraper | vca | setupsim | strategist | llm
User avatar
Mulch
Mulch
Gotta Go Fast
User avatar
User avatar
Mulch
Gotta Go Fast
Gotta Go Fast
Posts: 16940
Joined: April 23, 2017

Post Post #71 (ISO) » Sat Oct 07, 2017 10:32 am

Post by Mulch »

In post 70, Psyche wrote:
In post 69, Mulch wrote:See none that come without drawbacks
this will have bad effects and also good effects
Not worth it
Amrock#3784
User avatar
Psyche
Psyche
he/they
Survivor
User avatar
User avatar
Psyche
he/they
Survivor
Survivor
Posts: 10048
Joined: April 28, 2011
Pronoun: he/they

Post Post #72 (ISO) » Sat Oct 07, 2017 11:22 am

Post by Psyche »

ok
youtube playlist extracter | donbot | game scraper | vca | setupsim | strategist | llm
User avatar
LicketyQuickety
LicketyQuickety
Survivor
User avatar
User avatar
LicketyQuickety
Survivor
Survivor
Posts: 12785
Joined: May 14, 2015
Location: Where the moon and the sea meet.

Post Post #73 (ISO) » Sat Oct 07, 2017 7:18 pm

Post by LicketyQuickety »

I am with dave on this.
I was anything worse than you! Anything worse than you was I!

You was doided teh aposit_tisopa het dedoid saw em.
User avatar
LicketyQuickety
LicketyQuickety
Survivor
User avatar
User avatar
LicketyQuickety
Survivor
Survivor
Posts: 12785
Joined: May 14, 2015
Location: Where the moon and the sea meet.

Post Post #74 (ISO) » Sat Oct 07, 2017 7:29 pm

Post by LicketyQuickety »

Also, I want to encourage Psyche with his project.
I was anything worse than you! Anything worse than you was I!

You was doided teh aposit_tisopa het dedoid saw em.
Post Reply

Return to “Mafia Discussion”